ON DIFFUSION IN SOLIDS. J69 
observation of this kind. It has, however, been observed that mixtures 
of 80 to 93 per cent. of ammonium nitrate and 20 to 7 per cent. of 
potassium nitrate, which solidify to form a mixture of two saturated 
solid solutions, become turbid on cooling past 104°, and monoclinic 
crystals then appear, which gradually absorb both the original forms, 
the mass ultimately becoming homogeneous. In some cases the result- 
ing crystals are optically perfect, whilst in others traces of heterogeneity 
remain.*4° . 
The other case is more frequent. An instance described in the 
above memoir is that of mixtures of 67 per cent. of ammonium nitrate 
with 33 per cent. of cesium nitrate. These mixtures solidify to form 
homogeneous cubic crystals, which become rhombohedral at a lower 
temperature. On cooling still further, these crystals are resolved into 
_ two distinct kinds of crystals, respectively isomorphous with the two 
components. 
It will be sufficient to cite one other example among salts. Mixtures 
of sodium sulphate and potassium sulphate solidify in the form of 
homogeneous crystals throughout the whole range of composition, but 
at lower temperatures separation into the double salt, Na»SO,,K.SO,, 
and the excess of one or the other constituent takes place.'*7 
An interesting case of dissociation in, and crystallisation from, a 
solid solution has been observed in the case of o-nitrobenzaldehyde, 
which is conyerted into o-nitrosobenzoic acid in sunlight. The green 
unimolecular modification of the nitroso-compound is produced as long 
as it remains in a state of solid solution, but when this becomes satu- 
rated the nitroso-compound formed subsequently separates in crystals 
of the white bimolecular modification.14% 
_ It does not appear that diffusion has been observed in the case of 
zonal crystals of salts or organic compounds, although such crystals 
are readily prepared by allowing a salt to grow in a saturated solution 
of an isomorphous salt, and similar crystals have been frequently inves- 
tigated on account of the optical anomalies that they present. Very 
remarkable results have been obtained, however, by the examination 
of crystals which have absorbed a colouring matter, not isomorphous 
with the substance examined, during crystallisation. These crystals 
commonly present optical anomalies, and the distribution of the colour- 
ing matter is frequently not uniform, but is confined to certain zones 
or sectors. In some cases the anomaly is undoubtedly due to 
mechanical strain caused by the interposition of a foreign material in 
the mass of the crystal, as the same effect is observed in crystals built 
up of isomorphous salts,‘4® but in other cases the disturbance goes 
further than this. Without referring to the very voluminous literature 
of this subject, a few typical examples may be mentioned. Experiments 
with 33 inorganic salts and 26 colouring matters, mostly aniline dyes, 
M6 FF, Wallerant, Compt. rend., 1906, 142, 100. 
M7 R. Nacken, NV. Jahrb. Min. Beil. Bd., 1907, 24, 1. 
148 ©. A, Lobry de Bruyn and C. L. Jungius, Proc. k. Akad. Welensch. Amsterdam., 
1903, 5, 643. 
119 A review of the subject to 1891 is given by R. Brauns, ‘ Die optischen Anomalien 
der Krystalle ’ (Leipzig, 1891). 
1912. BB 
